Genetic deletion of aldose reductase protects the neonatal mouse retina from oxygen-induced retinopathy

Abstract

PURPOSE: Retinopathy of prematurity (ROP) is a common retinal disease that occurs in premature babies with initial vessel loss followed by vessel proliferation. Oxidative stress and retinal dysfunction has been reported. We previously showed that genetic deletion or pharmacological inhibition of aldose reductase (AR), a rate limiting enzyme in polyol pathway, prevented RGC loss and oxidative stress after retinal ischemia/reperfusion injury (Cheung et al Exp Eye Res 2007 85:608). Here, we assessed the effects of aldose reductase deletion on retinal injury induced by hyperoxic exposure using a mouse model of ROP. METHODS: Seven-day-old pups (P7) were exposed to 75% oxygen for 5 days and then returned to room air environment for 5 days. Vessel architecture and neuronal responses after hyperoxia were examined and compared between wild-type and AR-deficient retinae. Immunohistochemistry for immunoglobulin G (IgG), calretinin, glial fibrillary acidic protein (GFAP), nitrotyrosine (NT), inducible nitric oxide synthase (iNOS) were performed. RESULTS: At P12 (immediately after hyperoxia), AR-deficient retina displayed significantly smaller central avascular area (57.9±1.1% vs. 64.3±1.4% in wild-type retina, p<0.003). There was significant reduction in thickness of central INL and entire IPL in wild-type retina but not in AR-deficient retina (p<0.05). GFAP immunoreactivity was increased and the intensity was stronger in AR-deficient retina, but that for calretinin in wild-type and AR-deficient retinae was similar. At P17 (5 days after hyperoxia), central avascular area persisted in wild-type retina (24.8±1.7%). Again, AR-deficient retina showed a smaller area (16.1±1.2%, p<0.0005). Interestingly, blood vessels along GCL of wild-type retina displayed IgG extravasation after hyperoxia. Yet, this was much reduced in AR-deficient retina. There was absence of calretinin staining in INL together with seriously distorted strata in IPL of wild-type retina but not AR-deficient retina. Increased intense GFAP staining was seen in central-middle retina of wild-type but not AR-deficient mice. Increased NT immunoreactivity was found in INL of wild-type retina but this increase was less apparent in AR-deficient retina. Similar increase in iNOS immunostaining was observed in wild-type and AR-deficient retina. CONCLUSION: Our observations indicated that AR deletion reduced avascular area, IgG leakage, amacrine cell loss, and NT induction upon hyperoxia, suggesting a protective role of AR deficiency in ROP.